Andrea Canciani writes:
> > I may be wrong here, but seems like everyone has his own definition of what
> > is
> > considered to be "thread safe". Currently pixman appears to be thread safe
> > as
> > long as same pixman objects (pixman_image_t and the others) are not used
> > from
> > multipl
Siarhei Siamashka writes:
> > We probably want to implement DllMain on win32 to allocate/free TLS anyway.
>
> Yes, I think it makes sense to give this a try. The constructor attribute is
> supported since at least gcc 2.95, and it seems to be fine in clang 2.7 too.
> There's always a risk that
On Saturday 28 August 2010 12:21:52 Andrea Canciani wrote:
> On Fri, Aug 27, 2010 at 3:58 PM, Siarhei Siamashka
> > Having something like pixman_init()/pixman_cleanup() functions could
> > solve all the problems, but it's an API change. Other solutions (using
> > atexit(), using mutexes or atomic o
On Sat, Aug 28, 2010 at 5:34 PM, Siarhei Siamashka
wrote:
> On Saturday 28 August 2010 12:21:52 Andrea Canciani wrote:
>> On Fri, Aug 27, 2010 at 3:58 PM, Siarhei Siamashka wrote:
>> > This code which is setting a global implementation pointer is also not
>> > quite thread safe (though very unlike
On Sat, 2010-08-28 at 18:34 +0300, Siarhei Siamashka wrote:
> I may be wrong here, but seems like everyone has his own definition of what
> is
> considered to be "thread safe". Currently pixman appears to be thread safe as
> long as same pixman objects (pixman_image_t and the others) are not use
On Saturday 28 August 2010 12:21:52 Andrea Canciani wrote:
> On Fri, Aug 27, 2010 at 3:58 PM, Siarhei Siamashka wrote:
> > This code which is setting a global implementation pointer is also not
> > quite thread safe (though very unlikely to cause any practical problems
> > other than a bit bigger o
On Fri, Aug 27, 2010 at 3:58 PM, Siarhei Siamashka
wrote:
> On Thursday 26 August 2010 20:40:08 Andrea Canciani wrote:
>> Currently pixman allocates dynamically its implementations, but does
>> not free them.
>> This is not a true memory leak, since implementations are only
>> allocated once, but
On Thursday 26 August 2010 20:40:08 Andrea Canciani wrote:
> Currently pixman allocates dynamically its implementations, but does
> not free them.
> This is not a true memory leak, since implementations are only
> allocated once, but
> many debugging tools will report the asyimmetric malloc/free as
Currently pixman allocates dynamically its implementations, but does
not free them.
This is not a true memory leak, since implementations are only
allocated once, but
many debugging tools will report the asyimmetric malloc/free as a memory leak.
I wrote a patch that avoids allocating pixman_impleme